New steering system concept to minimize aerodynamic drag
 

So the basic idea is the total elimination of wheel skirting and minimizing the underbody gap with the wheels. The below photo is just a sample of the concept. There are other ideas that I am thinking about to implement in conjuction with this that I may design later on.

My main concern is the handling during the turning and wondering if it effectively narrows the front wheel base.

http://ecomodder.com/forum/attachmen...1&d=1237428062

What do you think?

So you're proposing to keep the wheels stiff and articulate the vehicle in the middle, like some construction trucks? Sounds a little complicated . Judging by your picture, if you start turning left, then something sticks out on your right. Maybe if the car body was flexible?

The steering mechanism that I have in mind is actually pretty simple but more on that later.

I think it would be easier and more effective to fair in the wheels and suspension of the regular steering types.

Hi,

Can you clarify what we are looking at, please?

You are looking at the birds eye view of rough car design shape. The entire front end turns in relation to the rear, the wheels in the front would remain stationary to themselves. The intent is to completely enclose the wheels for aerodynamic purposes.

I should have something better in a couple days that demonstrates the concept much better.

Hey Blue Bomber, it looks like even small angular movements of the steered section would expose a sharp protrusion on one side of the vehicle to the air flowing past - bad for a streamlined, attached flow.

Maybe you should consider flexible overlaps of the steered section skin over the trailing section, so that the protrusion can be smoothened over. You are a smart man so will get what I am vaguely describing without a diagram to explain... ;)

I realize that the edge would stick out, but it would only be significant at lower speeds when smaller turning radii's are used. At higher speeds the turning radius is much larger and thus requires less turn angle.

Also I think I can design it so there is no edge/gap on either side of the vehicle even when positioned to turned to full. I think it will require some geometry that is a bit more complicated then circle but I pretty sure it could work.

Not sure I completely follow.

Where is driver's seat located?
How many wheels? Located where? How are they oriented in each view?

Answering the above questions with block diagram additions to your drawing would be much appreciated.

TIA

I think it would be a complete semicircle that would do the trick, rather than a shallow arc currently depicted. But then it might be too inefficient in using space available.

Heres a more recent model that I was trying to make more accurate then the first screenshots. It still needs some modification before I actually cut into the model for the steering system, but it shows the general lay out of the vehicle.

This version the vehicle is about 10' 6" long, maybe 4' "6 wide (guessing on the width, I dont remember off hand, probably wider). Once I update the model tonight the nose will be wider, allowing the wheels to move further forward. This allows for a wider wheel base and more area that could accept the drivers seat. I'm also considered moving the vehicle to a two seat design because it would only require a very slight increase in the vehicles side.

In this design I really wasn't trying to copy the silhouette of the Aptera but as it evolved the shape made a lot of sense.

http://ecomodder.com/forum/attachmen...1&d=1237473277

This version doesn't include any enclosures for the wheels yet. Not going to bother designing them until I have everything else squared away for the general shape.

http://apgaylard.files.wordpress.com/2008/09/aptera.jpg

The rear wheel will probably get wrapped up in a similar way to the Aptera's.

Assuming the hash marks in your original are 12" apart - I added a few dimensional references, which make me think the car would have to be longer than 10.5 feet.

http://ecomodder.com/forum/attachmen...6&d=1237491587

Yellow Dot - If the rear of the car pivots off of this point (and the wheels don't steer independent of their respective body sections), the vehicle will steer decently. Otherwise the rear wheel is going to be dragged laterally across the pavement at each turn.

Dark Grey arrows - Overall tire radius is 12" - implying ~9" wheels???

Green & Light Blue arrows - Seat base is 15" x 15". I'd guess that even an anorexic, 5' 2" person would find that seat at least a little cramped. The butt of my overweight 5' 11" (180cm) frame needs/wants 20" wide x 18" deep (51cm x 46cm).

Red arrows - Sitting in my desk chair with height set so thighs are parallel with the floor and knees @ 90*, the distance from the intersection of seat base and seat back to the tips of my toes is 39". Sitting in the seat of my Ford Escort (knees @ ~150*, the distance is 44".

Why not just steer with the back wheel? Then the suspension on the two front ones could be much simpler since it wouldn't need to include steering components. The back one is already mostly out of the wind flow, so minor changes to accommodate any additional steering related parts shouldn't make much difference.

Admittedly one would have to drive such a beast a little differently. For instance, to leave a parked position by the curb the car would likely have to back out. Going in would be easy though, stuff the nose in going forward, crank the wheel over the other way hard, and the tail slides right over to the curb.

That is not a bad idea for doing it, and I might look into doing it that way as well.

The main complications I for see in doing it that way are these:

-Steering would probably have to be by wire
-Steering by wire would probably want an electric steering system
-Would probably require the vehicle to be frontwheel drive and thus need two motors
-Controlling a two motor setup is likely more expensive

Direct Benefits:

-Simplified body construction
-Possibly tighter turning radius (depends on much you can force the vehicle to turn without the rear wheel skipping)
-Indirectly allows for a more powerful drive system.

If I could keep that same system rear wheel drive it would be extremely maneuverable though. This could be done by either doing a hub motor setup (difficult to find one powerful enough that is actually on the market: 15kw+) or by enclosing the motor in with the rear wheel setup; a setup that requires strong considering in this application.

TestDrive:

Here is a marked up image:

http://ecomodder.com/forum/attachmen...1&d=1237497900

I haven't move any components around yet for the previous adjustments I noted I wanted to make in an earlier reply.

Also on my Mazda Protege from the front of my seat to the back of the pedals is ~24". I am only 5'8" but that is with my legs fully extended. So I assume if I design 6-10" of travel into the seat position that it could easily accommodate larger people. If I decide to move on with the front section steering there would be a footbox space cut out of the front section to allow the legs to clear into it some to provide more room. As long as the whole section didn't need to turn more then about 15-20 degrees this would not impede movement of the outer front section walls as they swing inward.

I definitely want to put some more thought into designing a rear swinging assembly similar to what pasadena_commut was suggesting. It would depend on me verifying that I could maintain rearwheel drive which I think would improve the steering in this situation.

As a side note: I am using Screenshots then pasting them into MS paint. I'd like to be posting these images cropped but I don't see an option for that in paint, any suggestions?

In Paint, you can change the size of the image. If you look close, there are handles on the edges, they're just small, and not very obvious. Or, you can go to Image -> Attributes, and change the height and width there.

I like the back wheel idea more as well. The first image looks like it would be difficult to implement, but I guess you have ideas on how to do that. It seems like it would make very wide turns though.

Y’know, I can see why you would want to eliminate the current steering system to improve aerodynamics, but did you ever stop to think that the current steering system that everybody uses exists for a reason?

Have you ever driven a wheelloader? That is essentially what this is – a high speed wheelloader. Wheeloaders steer by articulation the chassis just like this. Conestoga wagons used that system. For low speed (<10 MPH) construction/mining this is A-OK. The problem is that a wheelloader is nearly impossible to steer a straight course with. No big deal at slow speeds, but a life-threatening problem at higher speeds.

The five-bar steering linkage used today (called Ackerman steering after the guy who designed it) makes the outside tire describe a much wider radius than the inside tire. The whole system is nearly self-centering.

This was a technological problem that was solved a hundred years ago and aerodynamics are not a good enough reason to throw it out.

Wagon steering was thrown out for a reason: It didn’t work.

Why doncha just stay with Ackerman steering (we know it works) and widen out the vehicle to 60" but don't widen the front wheels so far they would need to protrude out of the body work at full lock. This was used back in the 30s and 40s on some high-dollar European cars. It seemed to work OK. with a 60" wide car you have to make the tail 160 inches long. The additional length would give you room for a second seat in tandem, making for a much more commercially viable car. A nice VW TDI diesel transaxle gives you a very efficient drive train and most of the rough spots have been worked out of that.

Those little guys on the side have been hiding from me =) Thanks for pointing them out!

I think I am pretty sold on rear wheel drive, I'm convinced I could integrate an motor to the steering assembly. I also think that doing it this way allows for two person seating, or possibly a 2+1 setup if I tuck the 1's legs between the front two seats without increase the vehicle's size.

I do need to figure out more about lofting properly before I spend more time on this in SolidWorks. You'll notice that the above drawings have the front nose cutoff (the rear is intentional, but I couldn't round it off if I wanted to atm). This has to do with the cross section sketches that I loft the surfaces across. I can't seem to figure out how to close it at the end without making the loft fail for one of the few possible errors I seem to come across:

"Unable to knit surfaces together"
"Self-intersecting geometry"
or a couple others I can't think of at the moment. Anyone that has some experience with Solidworks I'd love to find out how to fix this or how to change my approach at creating the solid body. I cannot telling you how many hours I've spent beating my head over that topic. FYI it's SolidWorks 2007

Big_Dave:

Thanks for the detailed response. One of my original primary concerns was how the wheel base decreases as the vehicle turns, probably resulting in poor vertical stability. I thought about your other concern already about it wanting to be uncontrollable. My solution to this is have the pivot point ahead of the wheel so it is self centering.

The quick and dirty sketch below illustrates the same thing, only with the rear wheel pivoting instead. The initial drawings from the beginning of the thread also had their pivot point ahead of the front wheels (actually the pivot point was about 2-3" in front of the car, the steering system was going to require an arclike track to accommodate for that).

With the sketch below what do you think about how the car would handle (ignore issues of going in reverse, reverse would only be at low speeds and this setup would have some sort of power steering)?

http://ecomodder.com/forum/attachmen...1&d=1237503599

Rear-wheel is even more unstable. Even the Ackerman mechanism could not stabilize rear wheel steering at speed. This time, use forklifts as an example. Very manueverable at low speed but they wander all over the place at speed.

Front-wheel steering with the Ackerman linkage rules for a reason.

I hate to see you beating your brains out re-inventing steering, when you could be bearing down on what otherwise looks like a good design.

Rear wheel drive is OK.

Yup, that's where drive by wire comes in (a-la stealth bomber or other "can't fly it without a computer" aircraft).

The rear wheel could also just caster, and manipulate the front wheel torque to steer, maybe.

So I thought about the physics of the rear wheel steering as per Big_dave's warnings and I realized he is almost certainly right. The rear wheel turning causes the rear end of the vehicle to travel in a direction opposite of that the front of the vehicle will be heading. This in turn torque the body and cause the real wheel to likely sputter across the road surface. Kinda like a semi trailer backing up and jacking at high speed.

So that brings us back to front wheel steering =)

What makes the idea of "wagon" steering work so poorly. The only issue I can see is with the varying frontal wheel base. The wheels spin independently from each still so I don't see issues with the effective differences between inner and outer wheels during a turn.

Thoughts?

Explain a bit more?

The stealth bomber doesn't have a vertical stabilizer, this reduces the radar signature, but also means the design is inherently unstable. It steers by using these little air brakes in the wing tips, open the left one a little and the plane yaws left, etc.. But keeping the plane pointing in the right direction requires more computations and reactions per second than humans are known for possesing. A human trying to steer it manually would undoubtedly tumble out of the sky. But with the help of a computer it does fly.

http://static.howstuffworks.com/gif/...h-bomber-1.jpg

Why not use a combination of ideas. Honda some years back had a prelude that was all wheel steering, albeit not much but it did improve cornering. Now take that idea and marry it with more modern computing and you get a front wheel with reduce turning radius to not pertrude out from the fenders and the rear wheels turning to compensate.

There are two effects to consider with rear wheel steering:

1. The direction the car turns is the opposite of the direction the wheel turns (this is like steering a boat with a rudder).

2. The steering is not self centering.

Computer controlled steering could compensate for the second effect. As others have pointed out, with the computer making adjustments, inherently unstable vehicles can be controlled such that, from the operator's perspective, they are stable. Cars are moving in this direction anyway, with computers applying brakes at various wheels during extreme handling situations to maintain vehicle stability.

The first issue cannot be compensated for. Consider what happens to a rear wheel steered vehicle which is parked directly next to a curb. It isn't (in general) possible to pull out forward, because to do so the rear wheel has to turn towards the curb, and the closer rear wheel will run right into the curb as the car moves forward. To pull out from the curb one must back out. Rear wheel steering does mean that in some instances the car just has to be driven differently.

If there is a single rear wheel, and it is both driven and steered, extreme steering angles become feasible. (Losing the drive linkage to another wheel permits this.). Imagine the car parked at the curb with cars very near each bumper. The rear wheel could be turned 90 degrees to move the back end directly away from the curb until it clears the rear car (which due to the narrow rear end, is not nearly so far as for a regular car), and then turned about halfway back towards center to pull the rest of the car out of the parking space.

Hmm, if the rear wheel can turn like that it might be possible to pull forward out of some parking spaces. Method: move as far back parallel to the curb as possible, crank the rear wheel straight towards the curb, drag the butt of the car to it, turn the rear wheel straight (parallel to the curb), move forward keeping the rear wheel parallel to the curb. This will push the front of the car to the left.

I suspect one could get used to driving a car with these properties, but it would be hard to drive this car one day, and a normal car another day.

Is the idea here to make something as complex as possible?

LOL Frank, was not my intention. This is exactly why I love posting about stuff like this on these boards. You can toss an idea around and get a lot of different insights to the benefits and complications of a scenario.

After all this talk I've decided to stick with a traditional steering system and will probably just design a part that attaches to the steering system that fills in the wheel wells at the body to smooth airflow.

I might consider making the wheels internal, but probably won't because it would require a larger wheel base/more frontal area. No sense killing the all the benefits of shielding the airflow from the wheels.

steering
 

The 1962 18-wheeler that GM researchers came up with utilized this configuration,although the curvature of the cut-line is reversed to make the leading edge of the trailing structure a prolate hemisphere rather than a parachute as appears in your graphic.No matter the position of the "lead" structure,the air is always striking a radiused "nose" of the trailing structure,with no flow separation.

Ya have the curve flipped would make more sense. What sort of results did they have with this sort of steering mechanism?

Most engineers find Kelly Johnson's dicta of KISS - Keep It Simple Stupid - is a good policy.

I like the idea of fully enclosing the front (steering) wheels with a shell that attaches to the steering link. Another possibility is take basjoos' roller operated skirt and make it cover the wheel more fully.

Your vehicle is fairly short and should be a snap to park.

Another point here is a low mass, high CG narrow three wheeler - unless there is some kind of banking mechanism included, there is a real possibility of this matchbox toppling/tumbling on its side while negotiating a tight turn at some speed.

Bomber, did you take this into account?

results
 

Blue Bomber Man,the comment was made in the context of the aerodynamics of the articulation.GM's work revolved around the forward structure being the tractor,and the following structure,the trailer.The tractor's steering was conventional.

smooth air
 

Have you seen the HONDA Dream II solar car? They put the steering wheels on lazy-Susan turntables,mated to double-wishbone suspension,where the wheels turned with their respective fairings.There is no conventional "wheel-house" as in conventional cars.The apparatus is rotated like an outboard motor on a boat.Perhaps a bit of toe-in is provided for "centering".Whatever you do,it's got to be very robust,unbreakable,zero-defect.Too much is riding on it!

The Honda Prelude's rear wheel steering had two modes of opperation.
1) Slow speed 0-15 or 20mph. The rear wheels turned opposite of the front. Not near the angle of the front wheels. This aided in maneuverability, shortening the turning radius. At higher speeds you have too much oversteer.

2) High speed 15 or 20 mph on up. Parallel to the front. Not near the angle of the front wheels. This induces an increased slip angle in the rear tires reducing oversteer.

Wagon steering.
Wagons have 2 rear wheels. Pivoting a front "axle" reduces the track width on the front axle leading toward tipping issues with a single rear wheel. Under normal driving conditions you would probably be ok. Evasive maneuvers would get real hairy real quick.

Personally I would go with a manual rack & pinion steering of the front wheels. Mount the wheel pants to the upright so they travel up & down as well as turn with the wheel. With a dual wishbone suspension with coil over shocks, you would not have any suspension components higher than the tire by the outside of the vehicle.

Unfortunately you cannot have the wheel skirts travel up and down with the suspension. If you did that it would collide with the body that is fixed to the frame. If I left enough gap there to allow for full suspension travel, it would avoid that issue, but it would negate much of the aerodynamic benefit.

I think I have a solution though. Very hard to explain without a visual, but the basic idea is there will be two sets of steering linkages. One goes to the wheel, the other to the wheel skirts. The wheel skirt linkage will not be hard fastened to the wheel skirt pivot point, rather it will be free to travel up and down. This might require some small sets of bearings or something to prevent binding but I think some variation of this concept will be an adequate solution.

I probably won't make a solidworks drawing of the idea until I have finalized the rest of the steering and front suspension. Which might require me to design the frame as well ;)

So it might be a few weeks before you see something for this idea.

On a different topic regarding 'wagon steering' Aerohead had mentioned the research done on it in the 60's regarding that and indicated that the pivot point was behind the wheels rather then ahead of the wheels as my earlier sketch showed. The problem I see with having the pivot point behind the wheels is the steering system will always pick a side and try to pull that way. So unless the system was design to prevent that the driver would always be forced to counter steer some.

If the pivot point is ahead of the wheels the system should want to self center. Of course this requires a less favorable body design that leaves an edge out of the body during a turn unless the width of the body is the same as the diameter of the turn.

BTW guys, if this vehicle ever comes to be, I am certain it will be because you all provided so much advice on different concepts and had different perspectives.

for a nice electric self contained drive unit, see "reach truck". The drive and steering units are in the same self contained area. Could be made to fit just about anything else, although they haven't ever been designed with too much power or speed (about 15 mph tops, IIRC, due to the relatively tiny, solid tire used.)

Some of them have had the drive and steering wheels placed on opposite sides, with the drive wheel acting like the front wheel of a shopping cart as far as lateral movement, but still being capable of driving, while the steering wheel freewheels, and can be powered to move laterally.

Very little lateral movement is required to steer at any speed when high caster angles turn toe-movement into camber... the problem with that, is that camber-based steering requires suspension correction to not kill tires... i.e. when you steer, the vehicles suspension needs to counteract centripetal force, so the body of the vehicle either remains mostly straight, or leans into the turn.. see "T-Rex 3-wheeled vehicle"

It steers from the front wheels, the suspension is set at a high camber angle, and allows the rear wheel to be a pivot point, which makes the chassis lean into corners. This makes it so that while the tires are losing contact patch due to over-cambered turning (leaning up on their sidewalls), the vehicle's suspension correction helps the vehicle "dig in" to the corner, so that it helps to correct the contact patch of the tire.

I remember an ad of some car showing how the rear wheels turned in the same direction as the front, aiding in changing lanes. I wondered that it would make turning in a full parking lot quite tricky. I wonder how Honda's Prelude switched from one mode to the other, say if the 15-20 mph were exceded with the wheels turned?

Steerable front and back wheels make the car more complex, but it would allow a smaller angle of turn, making it easier to cover the wheels with skirts.